Nauplius 13(1): 1-27, 2005 1 Peracarida: monophyly? relationships and evolutionary success Poore, G. C. B. Museum Victoria, GPO Box666E, Melbourne, Vic. 3001 Australia, e-mail: [email protected] Abstract Although the monophyly of the Peracarida has from time to time been questioned, this seems now to be no longer in doubt, Mysida excepted. A phylogenetic analysis of relationships between the orders based on parsimonious analysis of 92 morphological characters (Syncarida as outgroups) has discovered that: (1) Mysida and Lophogastrida are sister taxa and the most basal clade; (2) Isopoda and Amphipoda are sister taxa and derived; (3) Cumacea and Tanaidacea belong in the same clade as these two; (4) Spelaeogriphacea (Spelaeogriphus, Potiicoara and Mangkurtu) are a monophyletic taxon; (5) Mictacea {Mictocaris, Hirsutia and Thetispelecaris) are monophyletic and related to Spelaeogriphacea; (6) Thermosbaenacea are sister taxon to Spelaeogriphacea and Mictacea or more basal to a larger clade including these two orders; and (7) the hypothesis proposing an alternative ordinal structure, Cosinzeneacea and Bochusacea (for Spelaeogriphacea and Mictacea) is not supported. Acadiocaris and l^iaoningogriphus are supposed fossil spelaeogriphaceans but in fact display few of the characters defining that order. Previous attempts at elucidating peracarid relationships that were not based on a parsimonious treatment of characters produced widely divergent results, none similar to that proposed here. Others using smaller character suites treated parsimoniously are in many ways congruent with that proposed here. Two recent studies based on molecular data found that Mysida are more related to decapods or euphausiids and not to Lophogastrida, a result at odds with the morphological consensus. Spears et al.'s finding that Spelaeogriphacea and Amphipoda are sister taxa has no morphological support. The morphological cladogram relating peracarid orders suggests a succession of evolutionary events correlated with behavioral and ecological changes. Mysida and Lophogastrida are epibenthic swimming crustaceans with a carapace that survive by swarming. Other epibenthic orders with short carapaces (Thermosbaenacea, Spelaeogriphacea and Mictacea) are represented today by few species in relictual freshwater or deep-sea habitats. Only the cryptic, benthic Cumacea and Tanaidacea (with short carapace) and Isopoda and Amphipoda (lacking a carapace) are now hyperdiverse and superabundant in marine environments. Introduction Peracarids are crustaceans that usually brood their young in a marsupium formed by branches of the thoracic limbs. Some, isopods and amphipods, include dozens of families and are among the most common of all marine crustaceans. Representatives of these orders are found in fresh water and on land. Tanaidaceans and cumaceans are likewise predominandy marine but are slightiy less diverse. Spelaeogriphaceans, mictaceans and thermosbaenaceans on the other hand are rare (with few species) and confined to unusual habitats, caves, fresh water or the deep sea. Mysids and lophogastrids differ from these benthic taxa in being demersal or pelagic, and rarely non-marine. Poore, G. C. B. "Peracarida evolution 2 The relationships between the orders of Peracarida have long attracted interest and controversy. Even which orders to include or exclude has been debated. The debate has attracted the interest of many experienced crustacean biologists and has been summarised in greater or lesser detail before (e.g. Pires, 1987; Wagner, 1994; Schram and Hof, 1998; Hessler and Watling, 1999; Richter and Scholtz, 2001). Much of the uncertainty that has hindered wide acceptance of a common view has revolved around character interpretation, notably homologies in the carapace. Early contributions to the debate did not rely on what is now accepted as a standard methodology, parsimonious cladistic analysis of shared homologous characters, and some recent opinions do not depend on this methodology either. Approaches to the problem are various but have not been critically assessed in the light of modern methods. The need to re-examine peracarid relationships has also been stimulated by the disturbing findings from molecular genetics (Jarman et a/., 2000; Spears et ah, in press) that suggested that the composition of Peracarida was not as popularly accepted but, instead, more like views held in the nineteenth century. Morphological evidence has not to date provided a widely accepted phylogeny against which molecular evidence could be tested. Many authors have come to the problem of peracarid phylogeny from the bottom-up, as part of exploration of the Malacostraca, Crustacea or Arthropoda as a whole. My interest was sparked from the top-down, an investigation of the relationships between four living species of Spelaeogriphacea to each other and to two supposed spelaeogriphacean fossil taxa (Poore and Humphreys, 1998, 2003). The search for an outgroup for these was not satisfied by the contradictory views in the literature. The relationship between the species in three Gondwanan continents and two fossil taxa in Laurasian continents invites investigation of their phylogeny, in particular how this might explain their biogeography. The monophyly of most peracarid orders has not been disputed. Intermediate forms have not been reported even among fossil taxa (but see discussion on spelaeogriphaceans below). Amphipoda, Isopoda, Tanaidacea and Cumacea are each well defined by numerous unique autapomorphies. The same is true for Thermosbaenacea but they have switched from Peracarida (Monod, 1927) to their own separate superorder, Pancarida (Taramelli, 1954; Bowman and Abele, 1982) and back again (Wagner, 1994; Martin and Davis, 2001). Mysida and Lophogastrida (Richter, 2003) have always been thought of as monophyletic but whether they are two orders or one order, Mysidacea, has been debated. This has become an issue as Mysida has been shown from molecular evidence to be distant from Lophogastrida and other peracarids (Jarman et ai, 2000; Spears et a/., in press). The monophyly of Spelaeogriphacea, comprising three extant and two fossil genera (Poore and Humphreys, 1998; Shen et a/., 1998; Poore and Humphreys, 2003), and Mictacea, of two families and three genera (Bowman et a/., 1985), is not universally accepted. Gutu and Iliffe (1998) and Gutu (1998) pointed out significant differences between the two mictacean families, Mictocarididae and Hirsutiidae, and proposed new ordinal names for Hirsutiidae alone (Bochusacea) and Mictocarididae plus Spelaeogriphidae (Cosinzeneacea). This paper offers a critique and tests of competing hypotheses. It attempts once again to hypothesise relationships between peracarid orders. Relationships between the extant and fossil $sQ spelaeogriphacean genera, mictacean genera and the monophyletic peracarid orders are "^investigated. Three eucarid taxa, Order Euphausiacea and the suborders Caridea and Dendrobranchiata of Order Decapoda, are included because of their supposed affinities with mysids and lophogastrids. The data are morphological characters, essentially a compilation of those used previously but with additions. Nauplius 13(1): 1-27, 2005 . ^ •* Materials and Methods _____ Taxa chosen Phylogenefic (cladistic) methods were used to generate cladograms as hypotheses of the relationships of peracarid taxa. Characterisation of orders of undisputed monophyly was based on their hypothetical ground-pattern. Data was gathered direcdy from specimens in the extensive collections of Museum Victoria and complimented with generalities in the literature. Nouvel et al. (1999) and Tattersall and Tattersall (1951) were the principal sources for Mysida and Lophogastrida. Bacescu and Petrescu (1999) reviewed Cumacea, Gutu and Sieg (1999) Tanaidacea, Roman and Dalens (1999) Isopoda, Bellan-Santini (1999) Amphipoda and Wagner (1994) Thermosbaenacea. Views on the most plesiomorphic isopod have changed significantly recently and a phreatoicid-like or asellote-like ground pattern is chosen (Wagele, 1989; Brusca and Wilson, 1991; Brandt and Poore, 2003). Other orders appear more uniform and the the ground pattern less contentious. All spelaeogriphacean genera were included: three extant (monotypic except in one case) and the two supposed fossil species. Except for the Australian species, information relied on literature: Spelaeogriphus lepidops (Gordon, 1957, 1960), Potiicoara brasiliensis (Pires, 1987), and Mangkurtu mityula and M. kutjarra (Poore and Humphreys, 1998, 2003). Information on fossil species of purported spelaeogriphaceans came from published works: Acadiocaris novascotica (Copeland, 1957) in Carboniferous marine sediments in Canada (Copeland, 1957; Schram, 1974) and Uaoningogriphus quadnpartitus in lacustrine deposits of Jurassic age in China (Shen et al, 1998; Shen et al, 1999). Five species in three genera comprise the Mictacea: Mictocaris halope (Bowman and Iliffe, 1985), Hirsutia bathyalis and H. sandersetalia (Sanders et al, 1985; Just and Poore, 1988), and • Thetispelecaris remex and T. jurikago (Gutu and Iliffe, 1998; Gutu, 2001; Ohtsuka et al, 2002). Characterisation of the three genera relied on this literature and the one specimen of H. sandersetalia in Museum Victoria. Scoring of the three eucaridan taxa, Euphausiacea, Caridea and Dendrobranchiata, was based on McLaughlin (1980) supplemented by personal observations of specimens in Museum Victoria. Outgroups were chosen from Syncarida, a generalised member of Anaspidacea (close to Anaspides)